Proc. Nadl. Acad. Sci. USAVol. 86, pp. 8280-8283, November 1989Biochemistry

y-Monomethyl phosphate: A cap structure in spliceosomal U6 smallnuclear RNA

(mRNA splicing/RNA processing/RNA modification)

RAVINDER SINGH AND RAM REDDYBaylor College of Medicine, Department of Pharmacology, 1 Baylor Plaza, Houston, TX 77030

Communicated by Aaron J. Shatkin, July 28, 1989

ABSTRACT U6 small nuclear RNA (snRNA), a componentof eukaryotic spliceosomes, is required for splicing of nuclearpre-mRNAs. Whereas trimethylguanosine cap-containing U sn-RNAs are transcribed by RNA polymerase H, the U6 RNA istranscribed by RNA polymerase HI and contains a nonnucle-otide cap structure on its 5' end. We characterized the capstructure of human U6 snRNA and show that the y phosphateof the 5' guanosine triphosphate is methylated. The mobilities ofin vivo-modified y phosphate from the 5' end of HeLa U6 RNAwere identical to the synthetic monomethyl phosphate (CH3-O-P) in two-dimensional chromatography and two-dimensionalelectrophoresis. The cap structure ofU6 RNA is distinct from allother cap structures characterized thus far.

In eukaryotes, RNAs transcribed by RNA polymerase II,such as mRNAs, small nuclear RNAs (snRNAs), and mostviral RNAs, are blocked on their 5' terminus by a guanosinecap: m7GpppN in the case ofmRNAs and m2.2,7GpppN in thecase of U1-U5 snRNAs (reviewed in refs. 1-3). U2 RNA caphypermethylation requires the Sm-binding site (4). In con-trast, certain viral RNAs have protein covalently attached totheir 5' terminus (reviewed in ref. 1).U6 snRNA, a member of the U snRNA family, is a com-

ponent of eukaryotic spliceosomes (5-7) and is required forsplicing of nuclear pre-mRNAs (5, 6, 8, 9). In yeast Saccha-romyces cerevisiae, U6 RNA is encoded by a single-copy geneand is essential for cell viability (9). Based on the presence ofan intervening sequence in the U6 gene of fission yeast (10) atposition 52, corresponding to stem I in the U4-U6 RNAcomplex, it has been argued (11) that U6 RNA may be directlyinvolved in catalysis during pre-mRNA splicing. While tri-methylguanosine cap-containing U snRNAs are transcribedby RNA polymerase II (reviewed in ref. 12), U6 snRNA isunique in that it does not contain the Sm antigen binding siteand it is the only known capped RNA transcribed by RNApolymerase III (13-15). Since the observation that the 5' endof rat U6 snRNA is blocked by a cap structure, designatedXpppG (13), the structure ofX has not been characterized. Inthis study, we characterized the cap structure of human U6snRNA and provide evidence that the y phosphate of the 5'guanosine triphosphate is linked to a methyl group through anester bond.

MATERIALS AND METHODSIsolation of U6 Cap Core. HeLa cells were labeled with

[32P]phosphate (0.5 mCi/ml; 1 Ci = 37 GBq) in monolayercultures for 16 hr, and the U6 snRNA was obtained byfractionating whole HeLa cell 4-8S RNA on 10% polyacryl-amide gels. The labeled U6 RNA was sequentially digestedwith nuclease P1 and alkaline phosphatase. This digest was

used for electrophoresis on DEAE-cellulose paper at pH 3.5to obtain the U6 cap core. The structure of this cap core wasshown to be XpppG, where X was identified as a nonnucle-otide compound (13). The XpppG was digested with tobaccoacid pyrophosphatase (Promega), and the digest was frac-tionated by chromatography and electrophoresis. The la-beled pG and Pi, used as standards, were obtained by treating[a-32P]GTP with venom phosphodiesterase and alkalinephosphatase, respectively. The monomethyl [32P]phosphoricester (CH3-O-P) was prepared by incubating 1 ,uCi of[32P]orthophosphate in 10 ,ul of 10 mM Tris HCI (pH 8) with1 ml of methanol at 65°C for 16 hr. The conversion oforthophosphate to monomethyl phosphate was >90% asconfirmed by its comigration with unlabeled monomethylphosphate obtained from Sigma and release of Pi by alkalinephosphatase (data not shown). The formation of methylphosphate from orthophosphate and methanol has been re-ported (16).Chromatographic and Electrophoretic Analysis. Chroma-

tography on PEI-cellulose plates (Fig. 1 A, D, and G) wascarried out as described (17). The first dimension was devel-oped with water up to the origin, with 0.25 M acetic acid untilthe solvent migrated 10 cm, and with 0.88 M formic acid untilthe solvent front had migrated an additional 12 cm; thesecond-dimension solvent was 0.22 M Tris HCI (pH 8).Chromatography on cellulose plates (Fig. 1 B, E, and H) wasdone as described (18). The first-dimension solvent wasisobutyric acid/water/NH40H, 66:33:1 (vol/vol), and thesecond dimension solvent was 0.1 M sodium phosphatebuffer, pH 6.8/ammonium sulfate/1-propanol, 100:60:2 (vol/wt/vol). Electrophoresis was carried out as described (19)(Fig. 1 C, F, and I). The first and second dimensions were atpH 3.5 on cellulose acetate and DEAE-cellulose paper,respectively. The total amount of radioactivity analyzed inthe case of Fig. 1 A-F was approximately 1000 cpm each; inFig. 1 G-I, equal counts (1000 cpm each) from in vivo labeledcap components and synthetic standards were applied. Au-toradiography was done for 72 hr at -70°C by using XAR-5film and Lightning Plus screens.

RESULTSTo characterize the cap structure of U6 snRNA, the relativemobilities of various nucleotides and their modified counter-parts during chromatographic and electrophoretic separa-tions (13, 17, 18) were analyzed. The 2'-O-methylated nucle-otides, in relation to their unmethylated counterparts, mi-grate faster during electrophoresis on DEAE-cellulose paperat pH 3.5 and in two-dimensional chromatography on PEI-cellulose sheets (13, 17). In contrast, the 2'-O-methylatednucleotides migrate faster in the first dimension but slower inthe second dimension on cellulose plates (18). It has beenargued that these effects of hydrophobic modifications on the

Abbreviation: snRNA(s), small nuclear RNA(s).

8280

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Apr

il 1,

202

0

Proc. Natl. Acad. Sci. USA 86 (1989) 8281

mobilities of nucleotides probably result from an increase inboth mass and hydrophobicity, as well as a slight increase inpKb of the residues after methylation (18).The cap core obtained from the U6 RNA was digested with

tobacco acid pyrophosphatase and electrophoresed onDEAE-cellulose paper at pH 3.5. In addition to pG and Pi,one spot with electrophoretic mobility greater than Pi wasobserved (13). This modified y phosphate from the 32P-labeled U6 snRNA, designated Xp, was previously analyzedby chromatography and electrophoresis (13). The mobilities

IN VIVO :,

pCA

ofXp and XpppG during chromatography on cellulose plates,PEI-cellulose plates, and electrophoresis on DEAE-cellulosepaper (13, 15) suggested that X could be a methyl group.Therefore, we analyzed several phosphate derivatives, in-cluding CH3-O-P, by electrophoresis and chromatography.We used 32P-labeled pG, Pi, and CH3-O-P as syntheticstandards to compare with labeled Xp, Pi, and pG derivedfrom the HeLa cell U6 RNA cap core. Fig. 1 A and B showthe analyses of U6 cap components in two different two-dimensional chromatography systems. The mobility of Xp

IN VIVO

pm

IN VIVO " I

PA

C

Xp l

B

pA

pG pU Xp

12 Pi

SYNTHETIC 1":pC

I pA

pU,,_. H3C-0-P

G

ac/pA

pG Xp

1 _ H3C-O-P2 Pi

2x IN

DSYNTHETIC

:IpmA227G.,N'._pA

pG ,

2x

IN VIVO

SYNTHETIC)o pms2t7G

pA

122

x

E

pm7G

PC

PU

H3C-O-P

0P1

2 * pG

IN VIVOH +

pm7GI

PC ,.t

Xp +H3C-O-P

Xp

H3C-O-P

*1F

Pi

pG pU Pi2 PG

1i*ISI .

FIG. 1. Characterization of U6 cap structure by chromatography and electrophoresis. (A, D, and G) Two-dimensional chromatography onPEI-cellulose plates. (B, E, and H) Two-dimensional chromatography on cellulose plates. (C, F, and 1) Two-dimensional electrophoresis oncellulose acetate and DEAE-cellulose paper. (Top) Analyses of tobacco acid pyrophosphatase digestion products of HeLa cell U6 cap core.(Middle) Analyses of standard pG, Pi, and CH3-O-P. (Bottom) Analyses of a mixture of U6 snRNA cap components and standard pG, Pi, andCH3-O-P spotted together. The unlabeled pG, pA, pC, and pU were also included in the samples as internal standards and are indicated by brokencircles. Unlabeled pm7G and labeled pm3-227G were analyzed in the two-dimensional chromatography system of Silberklang et al. (18), and theirmobilities are indicated in B, E, and H. pG, Pi, and CH3-O-P were also analyzed separately, and their chromatographic and electrophoreticmobilities were the same as shown here. The X indicates the origin, and arrows show the first and second dimensions.

I pm7G

pC '-'Xp

pG (I I",pU Pi

2. pG

SYNTHETIC

Pi

FH3C-o-P

Pi *%

pG

11 Pi2X-

IN VIVO

SYNTHETIC

Biochemistry: Singh and Reddy

I2,2,7G311

I

PC

Dow

nloa

ded

by g

uest

on

Apr

il 1,

202

0

8282 Biochemistry: Singh and Reddy

o f3 aG

H 0 0 0 5' BaseNI

I 11 11 IH-C-0-P-O-P-0-P-O-CH2

I

3' 2'

OH

y Mono Methyl O U

Phosphate ester O=P-O-CH2 Base N2

OH

FIG. 2. Cap structure of human U6 snRNA. Diagrammaticrepresentation of the methylated y phosphate of the 5' nucleotide G(nucleotide N1) of human U6 snRNA. The 2', 3', and 5' represent thecarbon moieties of the ribose sugar.

from the U6 RNA (Fig. 1 A and B) was identical to that ofCH3-O-P (Fig. 1 D and E). Similarly, the mobility ofXp (Fig.1C) was the same as that of CH3-O-P (Fig. 1F) in two-dimensional electrophoresis. To eliminate experimental vari-ations during analysis on different chromatographic platesand electrophoresis on DEAE-cellulose paper, approxi-mately equal number of counts from in vivo labeled cap andsynthetic standards were mixed and analyzed by the samesystems. The Xp and CH3-O-P comigrated in all three sys-

tems (Fig. 1 G, H, and I), showing that Xp in the U6 RNAis a monomethyl phosphate ester. The structure of U6 RNAcap is shown in Fig. 2.

DISCUSSIONThe data presented in this study show that the y phosphate ofthe 5' end of human U6 snRNA is blocked by a monomethylphosphoric ester (Fig. 2). Any proposed cap structure for U6snRNA must account for the following properties attributed toXp from rat U6 snRNA (13): (i) it is a nonnucleotide; (ii) X islinked to Pi by an ester linkage since alkaline phosphatase canrelease Pi from Xp; (iii) it does not contain free amino group(s);and (iv) there are no periodate-oxidizable vicinal hydroxylgroups. The structure of y-monomethyl phosphate is consis-tent with all these observations. This and the fact that Xp andCH3-O-P comigrate in different chromatographic and electro-phoretic separations lend credence to Xp being a monomethylphosphoric ester involving the y phosphate of the initiationnucleotide of U6 snRNA. The 5' end of U6 snRNAs charac-terized earlier from rat (13), mouse (20), trypanosomes (21),Physarum (22), plants (23), and dinoflagellates (24) was shownto contain cap structure different from m22 '7G cap found inother U snRNAs. In addition, U6 snRNA is the most con-served of all the U snRNAs (9, 25); therefore, it is likely thatU6 snRNA from other species has the same cap structure asshown in Fig. 2.

Interestingly, U6 snRNA is the only known capped RNAtranscribed by pol III. Our earlier studies showed that thepromoter for the U6 gene is external (26); however, tran-scripts lacking sequences corresponding to the U6 snRNA donot get capped. For the purpose of capping, transcriptscontaining as few as 25 nucleotides corresponding to the 5'end of U6 snRNA were as good substrates as the full-lengthU6 snRNA (our unpublished results), indicating that 1-25nucleotides of U6 snRNA contain information necessary andsufficient for capping. These data indicate that informationfor the formation of U6 cap resides within the transcribedportion of the U6 gene and may explain why U6 is the onlyknown capped RNA transcribed by pol III. The 5' region of

U6 snRNA can potentially form a stem-loop structure (6, 9,13, 25), as validated by chemical modification data (20) aswell as phylogenetic comparison of known U6 sequences(25), indicating that the integrity of this stem-loop structureis important for the U6 snRNA function.The cap structure of mRNAs has been shown to enhance

the stability of mRNAs by protecting against 5' exonucle-olytic degradation and to increase translational efficiency byfacilitating the formation of the initiation complex (reviewedin refs. 1 and 2). Recent studies indicate that cap structureplays additional roles in mRNA biogenesis. These includetranscription initiation (27), generation of capped primersnecessary for influenza viral mRNA synthesis (28), pre-mRNA splicing (29, 30), and 3' processing of mRNAs (31,32). The virion RNAs of cowpea mosaic virus and threepicornaviruses (namely, poliovirus, encephalomyocarditisvirus, and foot and mouth disease virus) have been shown tocontain a covalently linked protein at the 5' end of the RNA(reviewed in ref. 1). In the case of poliovirus RNA, a specificprotein, VPg, with U residues attached to it, may serve as aprimer during RNA replication (33). The function(s) of tri-methyl guanosine cap structure in other U snRNAs and ofmethyl phosphate cap structure in U6 snRNA are not known.The methylation of the y phosphate may protect U6 snRNAfrom exonucleolytic degradation. In fact, while methylgua-nosine cap structures can be cleaved from capped RNAs byvenom phosphodiesterase, the U6 cap is resistant (13).This report on the biochemical identity of U6 cap repre-

sents the third category ofRNA cap structures, the other twobeing: (i) nucleotide caps found in mRNAs, U RNAs, andmany viral RNAs; and (ii) protein cap on certain viral RNAs(1-3). The characterization of the U6 cap structure willenable one to develop antibodies specific to U6 cap structure,which will be useful in studying the structure and function ofthe U6 cap and spliceosomal U6 ribonucleoprotein particle.Our observation that a short region within the U6 snRNAsequence contains the information for the capping of U6snRNA will help to understand the underlying principlesgoverning the formation of this cap structure.

We thank Harris Busch for encouragement and Kurt Randerath,Ramesh Gupta, Subrahmanyam Chirala, Dale Henning, and ThoruPederson for valuable discussions and comments on the manuscript.This investigation was supported by Grant GM38320 awarded by theDepartment of Health and Human Services.

1. Banerjee, A. K. (1980) Microbiol. Rev. 44, 175-205.2. Shatkin, A. J. (1985) Cell 40, 223-224.3. Busch, H., Reddy, R., Rothblum, L. & Choi, Y. C. (1982)

Annu. Rev. Biochem. 51, 617-654.4. Mattaj, 1. W. (1986) Cell 46, 905-911.5. Berget, S. M. & Robberson, B. L. (1986) Cell 46, 691-696.6. Black, D. L. & Steitz, J. A. (1986) Cell 46, 697-704.7. Grabowski, P. & Sharp, P. A. (1986) Science 233, 1294-1299.8. Steitz, J. A. (1988) Sci. Am. 258 (6), 56-61.9. Brow, D. A. & Guthrie, C. (1988) Nature (London) 334,

213-218.10. Tani, T. & Ohshima, Y. (1989) Nature (London) 337, 87-90.11. Brow, D. A. & Guthrie, C. (1989) Nature (London) 337, 14-15.12. Dahlberg, J. E. & Lund, E. (1988) in Small Nuclear Ribonu-

cleoprotein Particles, ed. Birnstiel, M. L. (Springer, Berlin),pp. 38-70.

13. Epstein, P., Reddy, R., Hening, D. & Busch, H. (1980) J. Biol.Chem. 255, 8901-8906.

14. Kunkel, G. R., Maser, R. L., Calvet, J. P. & Pederson, T.(1986) Proc. Natl. Acad. Sci. USA 83, 8575-8579.

15. Reddy, R., Henning, D., Das, G., Harless, M. & Wright, D.(1987) J. Biol. Chem. 262, 75-81.

16. Bieleski, R. L. & Young, R. E. (1%3) Anal. Biochem. 6, 54-68.17. Gupta, R. C., Randerath, E. & Randerath, K. (1976) Nucleic

Acids Res. 3, 2915-2921.18. Silberklang, M., Gillum, A. M. & RajBhandary, U. L. (1979)

Methods Enzymol. 59, 58-109.

Proc. Natl. Acad. Sci. USA 86 (1989)

Dow

nloa

ded

by g

uest

on

Apr

il 1,

202

0

Biochemistry: Singh and Reddy

19. Brownlee, G. G., Sanger, F. & Barrell, G. B. (1968) J. Mol.Biol. 34, 379-412.

20. Miura, K., Tsuda, S., Harada, F. & Ueda, T. (1983) NucleicAcids Res. 11, 5893-5901.

21. Mottram, J., Perry, K. L., Lizardi, P. M., Luhrmann, R.,Agabian, N. & Nelson, R. G. (1989) Mol. Cell. Biol. 9, 1212-1223.

22. Skinner, H. B. & Adams, D. S. (1987) Nucleic Acids Res. 15,371.

23. Kiss, T., Antal, M. & Solymosy, F. (1987) Nucleic Acids Res.15, 543-560.

24. Reddy, R., Spector, D., Henning, D., Liu, M.-H. & Busch, H.(1983) J. Biol. Chem. 258, 13965-13969.

25. Roiha, H., Shuster, E. O., Brow, D. A. & Guthrie, C. (1989)Gene 82, 137-144.

Proc. Nati. Acad. Sci. USA 86 (1989) 8283

26. Das, G., Henning, D., Wright, D. & Reddy, R. (1988) EMBOJ. 7, 503-512.

27. Shatkin, A. J. (1976) Cell 9, 645-653.28. Ulmanen, I., Broni, B. & Krug, R. M. (1983) J. Virol. 45, 27-35.29. Konarska, M. M., Padgett, R. A. & Sharp, P. A. (1984) Cell 38,

731-736.30. Edery, 1. & Sonenberg, N. (1985) Proc. Natl. Acad. Sci. USA

82, 7590-7594.31. Georgiev, O., Mous, J. & Birnstiel, M. L. (1984) Nucleic Acids

Res. 12, 8539-8551.32. Hart, R. P., McDevitt, M. A. & Nevins, J. R. (1985) Cell 43,

677-683.33. Flanegan, J. B., Pettersson, R. F., Ambros, V., Hewlett, M. J.

& Baltimore, D. (1977) Proc. Natl. Acad. Sci. USA 74, 961-965.

Dow

nloa

ded

by g

uest

on

Apr

il 1,

202

0